Synopsis The Bearskin Lake Air Services Ltd. Beechcraft A100 was on a regular scheduled flight, under visual flight rules, to Big Trout Lake Airport, Ontario, with nine passengers and a crew of two on board. The crew were flying the aircraft over a lake about four miles northwest of the airport for a landing on runway 14 when whiteout conditions were encountered. The aircraft descended in controlled flight into the frozen surface of the lake. The crew and several passengers sustained serious injuries. Rescuers from the local community reached the aircraft about two hours after the crash and all eleven survivors were rescued within four hours. The Board determined that, while the crew were manoeuvring the aircraft to land and attempting to maintain visual flying conditions in reduced visibility, their workload was such that they missed, or unknowingly discounted, critical information provided by the altimeters and vertical speed indicators. Contributing factors were the whiteout conditions and the crew's decision to fly a visual approach at low altitude over an area where visual cues were minimal and visibility was reduced. 1.0 Factual Information 1.1 History of the Flight The crew of the Beechcraft A100, C-GYQT, were conducting a scheduled flight from Sioux Lookout, Ontario, to Big Trout Lake, Ontario, as Bearskin (BLS)1 324. BLS 324 departed Sioux Lookout with nine passengers and a crew of two at 1133 central standard time (CST)2 and arrived in the vicinity of Big Trout Lake at approximately 1240. The captain briefed an instrument approach with a circling procedure to runway 14. On descent to the radio beacon, the crew reportedly encountered flight visibilities of one mile and were in visual contact with the ground. When the aircraft was less than five miles3 from the airport, the crew heard a position report from another aircraft completing an approach to the airport. To ensure safe separation from the aircraft ahead, the captain elected to fly under visual flight rules to the southwest of the airport. Air Traffic Services radar data was obtained from the Big Trout Lake radar source. The radar data indicated that the crew descended to about 150 feet above ground level (agl) approximately 4.5 miles from the end of the landing runway and maintained 200 to 300 feet agl for some 50 seconds prior to impact. Immediately prior to impact, the radar data indicated that the aircraft was about 3 1/2 miles from the runway at about 300 feet agl and descending at more than 1,200 feet per minute. Throughout the approach, the first officer flew the aircraft visually with occasional reference to his instruments, while the captain navigated and maintained terrain clearance by visual reference to the terrain and issued instructions to the first officer. At approximately five miles from the runway, the crew turned onto the extended centre line of the runway and received a radio report from the other aircraft of local visibilities of less than 1/2 mile. The aircraft flew inbound over a wide expanse of lake, and the captain lowered the flaps in preparation for landing. Shortly thereafter, the captain became concerned with the reducing visibility and looked in the Company Approach Procedures binder that he held on his lap. The captain was aware of the danger of whiteout and intended to revert to instrument flight if whiteout were encountered. He had not previously removed the approach chart for Big Trout Lake and clipped it in the approach chart holder because he had discovered that the binder rings were broken and taped shut when he had performed his initial approach briefing. He intended to provide new approach information to the first officer so that a full instrument approach could be initiated from their current position. When the captain looked up from the binder, he observed the altimeter indicating a descent through 1,000 feet above sea level (asl) and called to the first officer, Watch your altitude. Before a recovery could be initiated, the aircraft struck the frozen surface of the lake and bounced into the air. The captain initiated a recovery and then, concerned with the airworthiness of the aircraft, reduced power and attempted to land straight ahead. The aircraft crashed onto the frozen surface of the lake about 3/4 mile beyond the initial impact location. All passengers and crew survived the accident. However, the crew and several passengers sustained serious injuries. Rescuers from the local community reached the aircraft about two hours after the crash and all survivors were rescued within four hours. The more seriously injured were experiencing the effects of hypothermia when rescued. The accident occurred at 1248 CST, approximately three miles northwest of Big Trout Lake Airport, at latitude 5349'N, longitude 08953'W, at an elevation of 690 feet asl. 1.2 Injuries to Persons 1.3 Damage to Aircraft The aircraft was damaged beyond economical repair. 1.4 Other Damage The aircraft's main fuel tanks ruptured on impact, and the resulting fuel spill contaminated the snow in the area of the impact. 1.5 Personnel Information The crew was certified and qualified for the flight in accordance with the existing regulations. The company has followed a policy of having all of their pilots receive crew resource management (CRM) training. The captain had received CRM training, but the first officer, who was a relatively new hire, had not received the training. 1.6 Aircraft Information A review of available records indicated that the aircraft was equipped and maintained in accordance with existing regulations and approved procedures except for the following two discrepancies: i. the propeller synchrophaser had been removed from the aircraft because of unserviceability, but this action had not been logged in the Journey Log; and, ii. the radar altimeter (radalt) had been declared unserviceable [non-specific] in the Journey Log and subsequently logged in the Deferred Items Log section of the Aircraft Journey Log. However, the radalt had not been placarded unserviceable in the cockpit, as per the placarding requirements outlined in the company Maintenance Control Manual. The captain was aware of the unserviceability, and the radalt was not used during the approach. The radalt was found with its switch in the OFF position and the bug set at approximately 1,050 feet. 1.7 Meteorological Information Big Trout Lake is served by an Automated Weather Observation System (AWOS) and a voice generation system. The AWOS provides up-to-the-minute weather information to the pilot via radio. Meteorologists use AWOS observations in the production of aerodrome and area forecasts. The weather forecast for Big Trout Lake, issued at 0500 CST, and obtained by the crew prior to departure, indicated that for their time of arrival, there would be an occasional sky condition of 3,000 feet overcast with visibility greater than six miles in light snow. The weather forecast issued at 1100, just prior to their departure, was substantially the same. This forecast was amended at 1227 to an occasional sky condition of 1,000 feet overcast with visibility of one mile in light snow. However, the weather the flight encountered was localized and corresponded more closely to the visibility reports from the AWOS. The AWOS report received by the crew before they began their approach procedures was made at 1212 and indicated that the cloud was scattered at 900 and 2,100 feet agl and the visibility was 1.3 miles in light snow. The weather report available from the AWOS by radio broadcast as the crew began their approach was made at 1239 and indicated that the sky was clear below 10,000 feet with visibility of 9/10 of a mile in light snow. The temperature was -14 Celsius, and the altimeter setting was 29.96 inches of mercury. From the observations taken at the AWOS over the period of the occurrence, the visibility dropped from 9/10 of a mile at 1239 to 1/2 mile at 1300. The winds were from the south at 13 to 15 knots. The AWOS did not report any obscured or partially obscured ceiling in the precipitation. 1.8 Aids to Navigation The airport is served by a non-directional beacon (NDB) located approximately 0.6 nautical miles (nm) west of the runway; the frequency of the NDB is 328 kilohertz (KHz), which was tuned to the aircraft's automatic direction-finder (ADF) radio. An interim Transport Canada Approved Company Instrument Approach Procedure, dated 07 November 1994, provides a circling approach to the runway with a minimum descent altitude of 1,280 feet asl and an advisory visibility of two miles. The elevation of the airfield specified by the chart is 777 feet. Company instrument approach procedure charts are issued on 8- by 11-inch paper and are kept in three-ring binders by the company. The one binder provided in the occurrence aircraft was broken and the individual charts could not be easily removed by the crew. The captain placed the binder on his lap during the approach. The aircraft was equipped with a global positioning system (GPS), which was not an approved navigation aid for instrument flight rules (IFR) navigation, and was used by the crew as a backup navigation aid. 1.9 Communications The airport is located in uncontrolled airspace. An aerodrome traffic frequency (ATF) is designated within a 5 nm radius below 3,700 feet asl. Prior to descent into Big Trout Lake, the captain communicated on the company frequency with another Bearskin flight that had departed Big Trout Lake about 30 minutes earlier. The crew of the departing flight reported the weather as one mile in snow, with snow showers to the west. On arrival at Big Trout Lake, the captain communicated on the ATF with the crew of another company who were flying a right-hand visual approach to runway 14. 1.10 Aerodrome Information Big Trout Lake is a certified airport operated by the Ontario Government. The airport elevation listed in the Canada Flight Supplement dated 08 December 1994 and used by the occurrence crew is 738 feet. This elevation is 39 feet lower than the elevation specified on the Company Approach plate. The elevation specified on the Company Approach plate is based on a more recent survey and is the correct airport elevation for Big Trout Lake. Runway 14/32 is gravel and 3,900 feet long by 100 feet wide. A RAMP radar site is located approximately 1,700 feet west of the threshold of runway 14. 1.11 Flight Recorders The aircraft was not equipped with any flight recorders, nor was there any regulatory requirement for the aircraft to be so equipped. 1.12 Wreckage and Impact Information The initial impact occurred on the frozen surface of Big Trout Lake approximately 3.4 nm northwest of the airport on the extended centre line of the runway and approximately 3.1 nm northwest of the RAMP radar site. The belly luggage pod was destroyed on first impact and the contents were strewn onto the ice. The trail of baggage and luggage pod debris was oriented on a heading of approximately 140 magnetic (M). Snow, which fell and was compacted by the wind after the occurrence, obliterated any potential evidence of wing or propeller strikes in the initial impact area. The second (main) impact site was located on the ice surface of the lake approximately 2.7 nm northwest of the airport, slightly to the left of the extended centre line of the runway. The aircraft crashed while in a left bank, nose-low attitude, on a heading of approximately 071M. The nose and the underside of the fuselage were crushed and the fuselage was buckled in several places; however, the main shape of the cabin was retained and all windshields and windows maintained their integrity. The left wing attachment fittings were broken, and the left wing was buckled in several places. The top surface of the right wing appeared undamaged; however, the rear spar of the right wing was fractured, and the right inboard flap cables were stretched and internal sleeves displaced by the forward movement of the right wing during the impact. The rear fuselage was wrinkled and buckled in several areas and both rear horizontal stabilizer attachment brackets were broken, which allowed the stabilizer to move freely. The landing gear were in the retracted position, and the flaps were extended to the 30 position. All primary flight controls were attached, and the continuity of controls was verified. The wings and horizonal stabilizer had a slight accumulation of rime ice along the leading edge. The strips of rime ice were about 1 to 1.5 inches wide and up to 1/8 inch thick. In several places, the thin strips of rime ice had been covered with oil during the impact sequence. Both engines were displaced to the left by the downward and forward motion of the aircraft during the impact sequence. The propeller of the left engine was found approximately 150 feet behind the aircraft at the start of the second impact wreckage trail. All four blades of this propeller were still in the hub. Three of the four blades were severely curled along 3/4 of their respective lengths. Examination of the propeller mounting flange revealed that the mounting bolt threads had failed. All of the bolts were found in the propeller shaft mounting flange on the engine. The propeller and gearbox of the right-hand engine were detached as a unit and became trapped under the engine against the fuselage. The right propeller blades were severely curled in a manner similar to the left propeller blades. This evidence is consistent with the engines providing high power as reported by the crew. All seats, with the exception of the crew seats and the bench seat at the back of the aircraft, had been removed by rescuers. Three seat-belt attachment fittings had broken during the impact. The captain's seat pan was badly deformed and the mounting structure under the seat was broken. The seat track attachments maintained their integrity, but the vertical seat posts had failed as the cabin floor was driven upwards. The shoulder harnesses from both of the pilot seats had been cut by rescuers, leaving only three or four inches of shoulder harness material at the inertia reel. The bulkheads separating the crew from passengers had also been removed by rescue personnel. 1.13 Medical Information There was no evidence that incapacitation or physiological factors affected the crew's performance. 1.14 Fire There was no fire either before or after the occurrence. Emergency response services (ERS) were not available. 1.15 Survival Aspects 1.15.1 First Aid Kit Air Navigation Order (ANO) Series II, No. 11, the Aircraft First Aid Kit Order, requires that aircraft be equipped with first aid kits for the treatment of injuries likely to occur in flight or in minor accidents. One of the passengers was a nurse who treated the survivors using the aircraft first aid kit. She reported that the first aid kit was not adequate to deal with the injuries that the passengers received. In particular, there was a critical lack of pressure dressings to stop bleeding. The nurse improvised and used packed snow to control bleeding. 1.15.2 Survival Kit ANO Series V, No. 12, the Sparsely Settled Areas Order, specifies the type of emergency equipment that must be carried on aircraft operating in the sparsely settled region of Canada. The order permits exceptions for air carriers as authorized in the air carrier's operations manual. The Bearskin Lake Air Services Operations Manual, which was approved by Transport Canada, exempted this flight from the requirement to carry a winter sleeping bag for each person on board. All personal baggage and equipment was lost when the baggage pod was crushed during the initial impact. A small survival kit carried in the aircraft cabin contained some foil (space) blankets that were used to protect the most seriously injured. However, the nurse reported that hypothermia occurred in the two most seriously injured survivors who were immobile. 1.15.3 Emergency Locator Transmitter (ELT) ANO Series II, No. 17, the Emergency Locator Transmitter Order, requires that information concerning the location and operation of the emergency locator transmitter be made available to the passengers by means of a readily visible placard located in the aircraft's cabin or by other equivalent means. After the crash, the passengers had difficulty understanding the placard and could only follow its instructions with the aid of the seriously injured captain. The crew members and the passengers had difficulty ascertaining if the ELT had activated automatically during the crash. When they attempted to activate the ELT manually, they also had difficulty in determining if it was activated. Eventually, the crew turned on the aircraft electrical power to hear the ELT tone on the aircraft radio. The ELT did function correctly, and a search and rescue aircraft was dispatched and evacuated the passengers and crew to Winnipeg. The ELT is manually activated by a small toggle switch accessible through a push-in access panel approximately the size of a quarter coin. The access panel is located on the exterior of the aircraft fuselage just forward of the starboard stabilizer. The location is identified by a placard. The on/off positions of the toggle switch were difficult to identify when viewed through the small access panel. 1.16 Tests and Research 1.16.1 Altimeters The aircraft was equipped with three pressure altimeters and one radar altimeter. The aircraft was not equipped with either altimeter reference markers or an altitude alerting system. The radar altimeter was not functional and not used by the crew. The pressure altimeter on the captain's side of the instrument panel was an encoding altimeter and required aircraft electrical power to function. Encoding altimeters provide the aircraft's altitude to air traffic control (ATC) radars via the aircraft's transponder. The pressure altimeter on the co-pilot's side of the instrument panel was not an encoding altimeter and did not require aircraft power to operate. The co-pilot's altimeter used a different static source than the captain's altimeter. The third pressure altimeter was a blind encoding altimeter and was not intended for crew use during flight. The blind encoding altimeter used the same static source as the co-pilot's altimeter. A switch was used to select either the captain's encoding altimeter or the blind encoding altimeter to provide the aircraft's altitude to ATC radars via the transponder. The selector switch was positioned to provide the blind encoding altimeter altitude to ATC radars at the time of the occurrence. On the day following the accident, at approximately 1230, TSB investigators noted that the co-pilot's altimeter read 960 feet asl, and that the altimeter's sub-scale was set to 29.96 inches of mercury. The reading of the captain's altimeter was not considered valid since electrical power had been interrupted; however, the subscale was noted as set to 29.96 inches of mercury. The subscale settings of both altimeters corresponded to the altimeter setting transmitted by the Big Trout Lake AWOS at the time of the occurrence. At the time that TSB investigators observed the co-pilot's altimeter reading of 960 feet asl, the Big Trout Lake AWOS altimeter setting was 29.69 inches of mercury. An Atmospheric Environment Service (AES) specialist in barometry used the preceding data to compute the altitude that the altimeter would have indicated if the correct altimeter setting, 29.69 inches of mercury, had been set on the co-pilot's altimeter. The computed altitude that the co-pilot's altimeter would have indicated for the atmospheric conditions at the time of the TSB observation was 710 feet asl, approximately the elevation of the lake surface. The captain's and co-pilot's altimeters were tested and found to be working within required accuracy limits. At test altitudes of 500 and 1,000 feet, the captain's altimeter indicated 500 and 1,010 feet, and the co-pilot's altimeter indicated 500 and 990 feet. 1.16.2 Vertical Speed Indicators Both the captain's and co-pilot's vertical speed indicators were tested and found to be calibrated in accordance with national standards and functioning well. 1.16.3 Horizontal Stabilizer Trim Actuator The horizontal stabilizer trim actuator was tested and performed within acceptable parameters. 1.16.4 Passenger Seat-Belts Two seat-belt attachment fittings from the occurrence aircraft and two new, similar fittings from the air carrier were subjected to strength testing. The applicable testing standard for seat-belt fittings is Federal Aviation Administration (FAA) Technical Standard Order C22. This standard requires that the belt fittings be designed to withstand loads of at least 1,500 pounds. All four fittings were tested to destruction under conditions specified in the standard; the two fittings from the occurrence aircraft failed at 1,596 and 2,218 pounds, and the two new fittings failed at 1,972 and 2,180 pounds. 1.17 Organizational and Management Information The company has published standard operating procedures (SOPs) for the guidance of pilots of the Beech A100, Beech 99, and Metro aircraft. The amount of guidance provided in the Beech 99 and the Metro SOP with regard to approach procedures is significantly greater than that provided in the Beech A100 SOP. For example, the Beech 99 and Metro SOPs specify the approach briefing in detail and direct that both pilots have their approach charts displayed for the approach. The Beech A100 SOP does not have such specific guidance. The Beech 99 and the Metro SOPs discuss missed approach procedures and specify that, The captain may also elect to carry out the missed approach at any time he may feel it is unwise to continue. There is no discussion of the missed approach procedure in the Beech A100 SOP. Company SOPs are not mandatory and do not require approval by Transport Canada. (See Section 4.1.) 1.18 Additional Information 1.18.1 Whiteout The Transport Canada Aeronautical Information Publication (AIP), section Air 2.14(b), describes whiteout as an extremely hazardous visual flight condition. Whiteout occurs over an unbroken snow cover and beneath a uniformly overcast sky. Because the light is so diffused, the sky and terrain blend imperceptibly into one another, obliterating the horizon. The horizon, shadows, and clouds are not discernible, and sense of depth and orientation is lost; only very dark, nearby objects are discernible. In addition, the AIP indicates that whiteout can result from blowing snow and falling snow. The real hazard in a whiteout is that pilots do not suspect the phenomenon because they may be in clear air. In many whiteout accidents, pilots have flown into snow-covered surfaces unaware that they have been descending, and confident that they could see the ground. Consequently, when pilots encounter the whiteout conditions described above, or even suspect they are in such conditions, they should immediately climb if at low level, or level off and turn towards an area containing sharp terrain features. Pilots should not continue the flight unless they are prepared to cross the whiteout area using instruments, and have the skills to do so. 1.18.2 AWOS Visibility and Ceiling Transport Canada issued an Aviation Notice, dated 02 February 1995, in response to user concerns with the performance of some AWOS sensors. The Aviation Notice included an interim operational caution as follows: If aviation users encounter an AWOS report of clear below 10,000 feet (CLR BLO 100) when precipitation and reduced visibilities are also reported, it is a definite indication that this is an erroneous sky condition report. An analysis of minute-to-minute data at some sites indicates that this condition may persist, in some cases, for over an hour. Version 5.2 of the ceilometer algorithm has, in laboratory tests, almost eliminated false reports of this nature. This algorithm will be deployed as soon as possible after satisfactory field testing. Please consult NOTAM [Notice to Airmen] for the latest information on this subject. NOTAM 940461 for Big Trout Lake was contained in the weather package carried by the crew. This NOTAM stated the following: CYTL AWOS. If aviation users encounter an AWOS ceilometer report of CLR BLO 100 when reduced visibility, precipitation or surface obscuration are present, it is a definite indication that this is an erroneous sky condition report. In the event of any discrepancy between AWOS ceiling or visibility and that observed by a pilot in the vicinity, operations may be based on the ceiling, runway visibility or flight visibility as provided by pilot report. 1.18.3 AWOS Altimeter Setting The crew of the Bearskin flight that had departed Big Trout Lake about 30 minutes prior to the occurrence reported that they had received an inaccurate AWOS altimeter setting. They set their altimeters to the aerodrome elevation while they were on the ground, but initially could not remember the resulting subscale setting. Several days later, they recalled a subscale setting that would have resulted in an altitude reading of about 100 feet high. The co-pilot's altimeter from the aircraft used by this crew was tested. It was found to be functioning correctly and well within calibration limits. At test altitudes of 500 and 1,000 feet, this altimeter indicated 490 and 980 feet. The crew of another air carrier had landed minutes before the occurrence and did not report any discrepancy with the AWOS altimeter setting. Both crews used a field elevation of 738 feet, obtaining the elevation from data stored in their respective GPS data bases. The information stored in the GPS data base is based on the Canada Flight Supplement. The AWOS altimeter setting provided is the lower reading of two sensors. If a discrepancy of 0.04 inches of mercury exists between the dual sensors, the system will fail safe, and the altimeter setting will be missing from the AWOS weather report. An on-site calibration of the Big Trout Lake AWOS was conducted on 16 March 1995 by an Environment Canada technical service specialist. The dual pressure sensors were found to be well within calibration tolerances. 1.18.4 RAMP Radar Data The transponder of the occurrence aircraft was transmitting aircraft altitude information to the Big Trout Lake radar site. The last transmission was received at 1248, when the aircraft was approximately 3.1 nm from the radar site, and gave the aircraft's pressure altitude as 800 feet asl. A pressure altitude of 800 feet asl corresponds to an actual altitude of 710 feet asl under the barometric conditions that existed at the time of the occurrence. 1.18.5 Controlled Flight into Terrain Controlled flight into terrain (CFIT) accidents are those in which an aircraft, capable of being controlled and under the control of the crew, is flown into the ground, water, or obstacles with no prior awareness on the part of the crew of the impending disaster. Previous TSB reports have indicated that meteorological conditions were a significant factor in more than 50 per cent of these type of accidents, and that loss of situational awareness is a fundamental element in the cause of these types of accidents. 1.18.6 Situational Awareness To make correct decisions when flying an aircraft, the crew must have an adequate knowledge of what is happening around them, that is, situational awareness. Without situational awareness, the crew has no starting point for correct decision making; appropriate action cannot be taken unless the information on which the decisions are based is valid. 1.18.7 Information Processing A considerable body of research has been developed concerning information processing and decision making. This research has established that stress and high workload may lead to a narrowing of attention to the primary task at hand and to the most noticeable information source(s) as perceived by the pilot. The most noticeable information sources may not be the most objective, and under stressful conditions, important information may be missed or discounted without the awareness of the decision maker. Hence, the decision maker or pilot unknowingly may become less situationally aware, even though striving to retain a correct perception of the situation. 1.18.8 Ground Proximity Warning System A ground proximity warning system (GPWS) is designed to issue visual and aural warnings to the flight crew when their aircraft is too close to terrain, or when its terrain closure rate, rate of descent, or glideslope deviation becomes excessive. The warnings are based on GPWS internal logic, radar altimeter information, and the aircraft's configuration. GPWS has prevented many accidents where, until the warning was sounded, the pilots had been unaware that the aircraft was in danger because of its proximity to the ground or water. The occurrence aircraft was not equipped with a GPWS and none was required by the regulations. 1.18.9 Approach Chart Holders Company aircraft are equipped with approach chart holders. These holders are mounted on both control yokes and are designed to hold approach charts published in the Canada Air Pilot. 1.18.10 Visual Flight Rules Visual flight rules (VFR) weather minima are specified in ANO Series V, No. 3. When the occurrence aircraft was manoeuvring southwest of Big Trout Lake, it was operating in uncontrolled airspace below 700 feet agl. The weather minima specified for these conditions are not less than one mile visibility and clear of cloud.